ETTR and the future

I felt ETTR was indeed the best way to get clean images from the BMCC, but - and this probably illustrates my inexperience - when trying to balance lights on set, what a pain. And then explaining to the producer/director (who was grading the footage) that everything needs to be reduced by a stop or two...and seeing clips from the indie feature overexposed.

I never really got on with the BMCC or understood the praise. However it was ground breaking for the price. The BMPC suited me much better.

With the 4.6k sensor claimed to be 800 ISO I really hope if I go with it (my third BM cam), which seems likely, it's not designed to require you to ETTR.

I also felt the BMCC may have been designed this way for marketing reasons and the claimed DR.

Am I the only one? Are those of you used to film and exposing with a light meter perfectly happy with ETTR?

I'd like to expose the 4.6k sensor and get the image in-camera.

Lighting ratios would be the same relatively speaking whether ETTR or not.

There is no reason to have to instruct your colourist to drop the exposure - just set it in camera.

Expose for 400 and you are ETTR by 1 stop.
Expose for 200 and you are ETTR by 2 stops.

As demonstrated in this test - it shouldnt affect the final image by setting it to either 200 or 400 in camera for raw - rather than 800 and then adjusting it to the correct level in post:




For ProRes or DNxHD - you would want to do the pulling in camera before compression anyway - so again, set it to the correct level in camera.



As an added bonus - it makes it much easier to monitor on the LCD.

There is no reason I can think of to film in an ETTR process and still have the camera set to 800.

With the 4.6k sensor claimed to be 800 ISO I really hope if I go with it (my third BM cam), which seems likely, it's not designed to require you to ETTR.

I also felt the BMCC may have been designed this way for marketing reasons and the claimed DR.

The camera is not "designed to require you to ETTR" - it just has a sufficiently high bit depth towards the top end to allow it to be possible in the first place. For the sake of argument - lets assume that the camera does indeed capture a full 13 stops of DR - and you want to film which happens to contain that level of difference in the light levels. You would need to shoot this at 800 - as this is where the signal is not boosted or attenuated digitally - giving you a completely captured scene.

In reality - many many scenes either A) don't contain that massive range or B) the parts of a shot which are above a certain point are not important to capture, such as the sun or a specular highlight.
So in these situations - because the camera has such a high bit depth (22bits before DSP to 16) - you can decide to either:
A) keep it at 800 and not fully use the entire range
B) Shift your exposure slightly to more fully saturate the sensor, giving a better signal to noise ratio and thus cleaner image.

This option is not a bad thing, or a design - it is just possible because of how the camera works.


Many cameras effectively allow for it to different degrees - Even the 5D Mkii with its ASA 50 - that is ETTR by 1 stop. It digitally reduces the image from 100 by 1 stop to achieve it. It gives you less range but a cleaner image.


Its a choice not a requirement.

Having the option is never a bad thing.

Would you rather have 800 and not be able to get a cleaner image at lower values????
I cannot fathom how this is somehow considered a bad thing.

1. The BM camera manual clearly says that the exposure should be maximally possible in condition that the overexposure does not take place. And it's a policy of BM that everything that can be moved to postprocessing shall be moved there.

In my experience Bmcc there is no need to ettr, i like a little organic noise in the background - it make the picture more filmlike. If someone struggle with right exposure I advise monitor with the right tools. I use samurai blade as monitor and recorder. And rgb parade and false color is saving me a lot of my time and nerves. If you expose with false color and keep your skintones with right area then image look clean and stunning. Rgb parade helps a lot when you want to set right color balance for monitoring and when shooting proress. I use samurai blade as backup recorder in proress - it alwaye better to be prepared. As one of my favourite movie characters once said - **** happens in blade is log to rec709 preview - it is not perfect but it gives you a hint of what the image will look after appling Lut.

i like a little organic noise in the background - it make the picture more filmlike.

I quite agree! For my own projects I prefer to shoot this way - but for clients, I tend to shoot at 200 or 400 - they expect it to be very very clean, and I prefer not to use NR.

You get fewer headaches when working on narrative projects if you use a light meter.

Elsewhere on this forum, there are discussions about how wide shots and close ups in the same location can vary widely, creating problems for ETTR. The best way to have consistency between setups is to meter.

You may find that your camera has a sweet spot that's a stop or two brighter than the meter advises. So you adjust the meter, or just change the stop in your head. A meter also lets you check ratios easily, and you can measure the light without moving the camera.

You get fewer headaches when working on narrative projects if you use a light meter.

Elsewhere on this forum, there are discussions about how wide shots and close ups in the same location can vary widely, creating problems for ETTR. The best way to have consistency between setups is to meter.

You may find that your camera has a sweet spot that's a stop or two brighter than the meter advises. So you adjust the meter, or just change the stop in your head. A meter also lets you check ratios easily, and you can measure the light without moving the camera.

Yea exactly - a consistent ETTR wont necessarily get you the cleanest image, but it makes more sense to getting consistent exposure and results. The best of all worlds would be to meter for 400 and shoot 400. That way you are ETTR by 1 stop but also don't have to manually adjust each shot in post.

Everyone seems so confident in this technique, but if you meter and set the camera to 400, how are you aware of clipping as in reality the camera is always 800? And you can't use a waveform monitor, right?

You get fewer headaches when working on narrative projects if you use a light meter.

Elsewhere on this forum, there are discussions about how wide shots and close ups in the same location can vary widely, creating problems for ETTR. The best way to have consistency between setups is to meter.

You may find that your camera has a sweet spot that's a stop or two brighter than the meter advises. So you adjust the meter, or just change the stop in your head. A meter also lets you check ratios easily, and you can measure the light without moving the camera.

Yea exactly - a consistent ETTR wont necessarily get you the cleanest image, but it makes more sense to getting consistent exposure and results. The best of all worlds would be to meter for 400 and shoot 400. That way you are ETTR by 1 stop but also don't have to manually adjust each shot in post.

Everyone seems so confident in this technique, but if you meter and set the camera to 400, how are you aware of clipping as in reality the camera is always 800? And you can't use a waveform monitor, right?

You get fewer headaches when working on narrative projects if you use a light meter.

Elsewhere on this forum, there are discussions about how wide shots and close ups in the same location can vary widely, creating problems for ETTR. The best way to have consistency between setups is to meter.

You may find that your camera has a sweet spot that's a stop or two brighter than the meter advises. So you adjust the meter, or just change the stop in your head. A meter also lets you check ratios easily, and you can measure the light without moving the camera.

Yea exactly - a consistent ETTR wont necessarily get you the cleanest image, but it makes more sense to getting consistent exposure and results. The best of all worlds would be to meter for 400 and shoot 400. That way you are ETTR by 1 stop but also don't have to manually adjust each shot in post.

When the zebras say 100% - its clipped.

People are confident in the method because it works really well and it is very easy.

Personally, I've found that setting my camera (BMCC) to ASA 400 and exposing "properly" gives the best results. With 400, you are already exposing to the right one stop.

I've also found that at 800, it's noisier and there's more colored noise, which I'm not a fan of. I also find that FPN shows up at 800 more often than 400. 400 seems to be the best ASA setting for me and I've found that I can even underexpose a shot and I still get great results. I don't know what it is, but maybe something in the 400 curve hides the noise better than 800.

You may very well have to bring the exposure down in post, but with the new Resolve 12 coming next month, there is now an ISO setting. So if you shoot 800 and overexpose by one stop, maybe tell your producer/colorist to knock it down one stop in Resolve by switching it to ISO 400.

For on set monitoring, if you have a monitor that supports it, use a LUT.

This approach wasn't taken when shooting with the BMPC though was it?

This approach wasn't taken when shooting with the BMPC though was it?

What approach?

Wasn't taken?


The production camera is a different camera and behaves differently. I don't see what point you are making?

The camera wasn't deliberately designed to force the user to expose to the right - nor was the production camera designed not to be done so. It is just how they work with their sensors.

With the cinema camera you have the option to utilize the unused or unneeded range of the camera and get a cleaner image if you want. You don't have to do this if you don't want to.

You know, exposing to the right. People didn't tend to take that approach with the BMPC.

I'm not trying to be critical, I'm just trying to understand a situation that I found confusing and making the point that I hope the new sensor is more like the BMPC, but sensitive to 800 ISO rather than the BMPC's 400.

I'm not trying to be critical, I'm just trying to understand a situation that I found confusing and making the point that I hope the new sensor is more like the BMPC, but sensitive to 800 ISO rather than the BMPC's 400.

This is because the Production camera had a lower dynamic range - so finding a situation where you can expose to the right and not clip any detail you want to keep is less common - HOWEVER - you are incorrect in what you say, because anybody shooting at 200 on the production camera is exposing to the right by 1 stop.

It is not a bad thing that a sensor can be exposed so close to clipping and have detail brought back - many cameras lose detail even before they clip - and im not even talking about single colour channels clipping.

Being able to do this has no drawbacks and only benefits - so why you don't want it for the new camera is what is confusing.

I'm not trying to be critical, I'm just trying to understand a situation that I found confusing and making the point that I hope the new sensor is more like the BMPC, but sensitive to 800 ISO rather than the BMPC's 400.

I get what you're saying. With the BMCC, the optimal shooting ISO is 400 (personal preference for me). You would like the new 4.6K sensor to perform similarly to the current BMCC 2.5k sensor but with an optimal ISO (i.e. similar performance to the BMCC 2.5K ISO of 400) of 800.

The Production Camera/current URSA is a different monster. With the BMCC at 400 ISO, I find that I can underexpose and bring it up without much noise that's unusable. The noise that's there is more like film grain.

When I bring up the URSA footage, if I bring it up the same way I do with the BMCC footage, I get FPN. What's usable or not is personal preference. To me, I have to be more careful about how much I underexpose, otherwise when I bring up areas of the image, FPN would show up. So in this case, exposing to the right is probably even more important because you can't underexpose as much on the URSA at 400 compared to the BMCC2.5K at 400.

High contrast shots, like filming someone against a sky, need to be carefully exposed on the URSA. Filling in the person's face just right or letting the sky blow out are more common. With the BMCC, I find that I could protect the sky a bit more and underexpose the actor a bit more than I can on the Ursa. Then, in post, I could bring up the BMCC footage. The Ursa is less felxible here.

I also find that sometimes underexposing the BMCC at 400 can give a nice look because it adds more of the grain-like noise.

So, all this to say that if you shoot at 400 on your BMCC, you should get better results. Check for yourself to see if it gives you a look you like, as it's largely personal preference. As for the 4.6K sensor, I've been wondering myself if 800 will give me similar performance to the BMCC2.5K's 400 ISO.

We'll just have to wait and see...

Cheers Paul. I think a lot of what I have said comes from the fact that I have worked as a TV cameraman for 15 years - almost all multi camera, so my exposure has been controlled for me. I'm trying to take my TV drama experience and learn more about exposure/lighting for film production.

What you describe in terms of exposure choices between letting the sky go or filling in the subject's face seems very familiar from my work. Also I think because of my background, I see the grain of the BMCC, that people like, as noise - I'm used to a cleaner image and that's why I took the advice to ETTR.

Once I got the BMPC I always worked at the native ISO and used a waveform monitor in my EVF to adjust exposure, taking the exposure up till the waveform flattened out and reducing the exposure until the shape returned when shooting wide exteriors. Shooting tighter, I'd expose for skin. This way, even in film dynamic range I would work with an image that was being displayed as it was being recorded and monitored on the waveform. Once in Resolve, the image and waveform would look the same - i.e. I shoot it right in-camera. The only thing required in post would be a small tweak of saturation, contrast and to sit the black level down a bit, unless I wanted more of a look. If you know where I'm coming from, you probably understand my need to get it right in-camera and monitor it as such. But if someone says you can achieve so much more with a different way of thinking, I'm all for it. I have my BMCC up for sale, but I must have a go once more at what you suggest shooting at 400 ISO

I still don't know what problem you have. Shoot at 400 on the cinema camera, Meter for it, use the zebras to see when it clips. It's really no different than what you or Paul had described.

Are you saying you wish the native iso was less noisy?
Are you saying you wish the camera had a lower native iso?

Cheers Paul. I think a lot of what I have said comes from the fact that I have worked as a TV cameraman for 15 years - almost all multi camera, so my exposure has been controlled for me. I'm trying to take my TV drama experience and learn more about exposure/lighting for film production.

What you describe in terms of exposure choices between letting the sky go or filling in the subject's face seems very familiar from my work. Also I think because of my background, I see the grain of the BMCC, that people like, as noise - I'm used to a cleaner image and that's why I took the advice to ETTR.

Once I got the BMPC I always worked at the native ISO and used a waveform monitor in my EVF to adjust exposure, taking the exposure up till the waveform flattened out and reducing the exposure until the shape returned when shooting exteriors. This way, even in film dynamic range I would work with an image that was being displayed as it was being recorded and monitored on the waveform. Once in Resolve, the image and waveform would look the same - i.e. I shoot it right in-camera. The only thing required in post would be a small tweak of saturation, contrast and to sit the black level down a bit, unless I wanted more of a look. If you know where I'm coming from, you probably understand my need to get it right in-camera and monitor it as such. But if someone says you can achieve so much more with a different way of thinking, I'm all for it. I have my BMCC up for sale, but I must have a go once more at what you suggest shooting at 400 ISO

Yeah shooting at ISO400 for me gives me a clean image when "eyeballing it" so it looks correct on the screen. I had the same problem with 800 that you did when I first got the camera. I thought it was really noisy. Now, not so much since I locked it at 400. I literally never move it. I just had a shoot last week where we used it for 2 days and I only shot at 800 for ONE shot, carefully watching for noise.

You definitely CAN get a good image from 800, but you have to make sure you have enough light. If you underexpose at 800, there's definitely more noise.

I never use 200. The image seems more dense in a bad way.

Try 400, expose properly so it looks right, and let us know what you think.

Here are some grabs with a quick one light that were shot outside at 400. Very little noise. In fact, I added sharpening so the noise that's there is pronounced. I will be adding more grain in the grade.










This shot is 800ASA

It's not just the BMCC that benefits for ETTR. Many cameras that shoot in LOG space benefit for ETTR.

ETTR is simple a way of getting the most out of the highlight range of the sensor and colour space.

Generally TV cameras shoot in REC709 so they have less headroom on the highlights and so there is no reason to even think about ETTR. The BMCC is not a TV camera...

Sometimes I feel like I understand ETTR but then so many people have conflicting information on here. I trust Tom because his footage is always beautiful, but dang it Tom, you have me baffled again.

I thought ETTR was related to the scene needs based on the meter reading. Let's say the camera and the meter are both set at ISO/ASA 800 and the scene reads at f/4. I thought exposing to the right by a single stop would be changing the aperture to f/2.8. That's adding a stop of light, thus exposing to the right of the scale. Changing the ISO to 400 would be taking away a stop of light (although I know with raw that 200 and 400 are just metadata and it's still shooting 800 but bringing the image visually down by 1-2 stops without modifying the image).

Isn't exposing to the right adding a stop or 2 to the scene based on how the scene reads? So to shoot ETTR by 2 stops consistently would be to set the camera to 800 and meter at 200 right? With raw, the camera can be set at 200 since the sensor is still reading at 800. But this isn't the case with ProRes.

Jim, with all that experience with TV drama, we can understand why you would look upon sensor noise as 'noise' rather than film grain. Video often tries to get a very clean image in a very controlled setting. A lot of people here are going for a different aesthetic as the goal is that organic film or cinematic look often with a wider range of exposure values perhaps than a typical TV drama (not talking Game of Thrones or the Vikings). Just an observation.


Sent from my iPad using Tapatalk

I'm away from home, so only viewed Paul's images on my phone but they look really nice. I'll look again on my Mac at home. I realise my method of exposure is based on rec709 not raw so despite owning the BMCC for two or so years I'll have to try some more.

Okay, here is my ct2:

As Adam said, the BMCC is not a TV camera, so you might have to relearn your habits a bit.

What I do is, first I set my ratios via false color. I use a TVLogic monitor with variable false color, so I can set my false color skintones/middle gray, at any IRE I want. Than I saturate the sensor (i.e. ETTR) as much as possible - usually at 800iso, if it is a lowlight scene, I might get higher. Than I lower the exposure just a hair, to make sure, that no single channel is clipping, cause zebras only come up, when all 3 channels are clipping.

When you need better monitoring for nervous clients/directors, get a monitor that accepts a LUT.

Jim I think only 50% of your problem is in the exposure, but 50% in your post workflow.
I guess you still think too much TV

I suspect you're right Frank

But what about Prores?

With raw i set my camera to 400 and expose to it. That way you ETTR 1 stop because the native iso is 800. And i check my false color for ratio's and IRE levels. And set the zebra's to 90, to check when things are clipping. It's just ''metadata'' with raw.

In Prores i use 800, when using 400 with prores, the image cleaner but the DR is ''smaller'' right? Even more with 200.

I didn't tested this properly but you are defintly losing some detail in highlights. Please correct me if I'm wrong.

Sometimes I feel like I understand ETTR but then so many people have conflicting information on here. I trust Tom because his footage is always beautiful, but dang it Tom, you have me baffled again.

I thought ETTR was related to the scene needs based on the meter reading. Let's say the camera and the meter are both set at ISO/ASA 800 and the scene reads at f/4. I thought exposing to the right by a single stop would be changing the aperture to f/2.8. That's adding a stop of light, thus exposing to the right of the scale. Changing the ISO to 400 would be taking away a stop of light (although I know with raw that 200 and 400 are just metadata and it's still shooting 800 but bringing the image visually down by 1-2 stops without modifying the image).

Isn't exposing to the right adding a stop or 2 to the scene based on how the scene reads? So to shoot ETTR by 2 stops consistently would be to set the camera to 800 and meter at 200 right? With raw, the camera can be set at 200 since the sensor is still reading at 800. But this isn't the case with ProRes.

When shooting a lower ISO in RAW you are effectively ETTR because you are feeding the sensor more light.

Eg: You have a scene that when exposed at 800ISO need an aperture of f5.6.

If you drop the ISO to 400 you have to open the aperture by one stop. So you are feeding the sensor an extra stop of light.

If you drop the ISO to 200 you have to open the aperture by two stops. So you are feeding the sensor an extra 2 stop of light.

In Prores i use 800, when using 400 with prores, the image cleaner but the DR is ''smaller'' right? Even more with 200.

I didn't tested this properly but you are defintly losing some detail in highlights. Please correct me if I'm wrong.

Correct but it's not that much smaller. Shooting at 400 in ProRes is usually a happy middle ground.

Yes, given my experience so far with the BMCC, I wish it were less noisy, but as pointed out I may be able to do better in this regard.

Do I wish the sensor was more sensitive? Yes, I'd like to expose for 800ISO.

I still don't know what problem you have. Shoot at 400 on the cinema camera, Meter for it, use the zebras to see when it clips. It's really no different than what you or Paul had described.

Are you saying you wish the native iso was less noisy?
Are you saying you wish the camera had a lower native iso?

Yes, given my experience so far with the BMCC, I wish it were less noisy, but as pointed out I may be able to do better in this regard.

Do I wish the sensor was more sensitive? Yes, I'd like to expose for 800ISO.

I still don't know what problem you have. Shoot at 400 on the cinema camera, Meter for it, use the zebras to see when it clips. It's really no different than what you or Paul had described.

Are you saying you wish the native iso was less noisy?
Are you saying you wish the camera had a lower native iso?

Ok - then really this has little to do with being able to expose to the right or not. You just want a cleaner image at asa 800.

If you nail exposure at 800 - with a touch of noise reduction it can look very clean.

Example:

Raw 800 - No Noise Reduction

http://www.tommajerski.com/publicimages/800_noNR.jpg

Raw 800 - Moderate Noise Reduction

http://www.tommajerski.com/publicimages/800_NR.jpg

The first thing I shot with the Pocket was a candle in a dark room with a mirror 2 metres behind, realised it made more sense to shoot at 400 ISO (BMPC) as I quickly realised, not all colours expose to the same level.

Given the camera has a massive dynamic range I saw the difference as risk management and not all that significant to the final outcome.

When I have my Alphatron as part of what I shoot, then I just expose to suit the waveforms and false colours and yeah I might be nuts, but I love shooting and love colour grading even more so get a kick from being able to stretch, shrink and crease the image....am the kinda guy that loves purple skies (;->>

Sure I might lose some of the top and some of the bottom, but every shoot is well within what I need.

As for ETTR, is not something I have ever thought of as I just expose to suit my mood.

Cheers Tom, I'll look over those when I get home at the weekend

Yes, given my experience so far with the BMCC, I wish it were less noisy, but as pointed out I may be able to do better in this regard.

Do I wish the sensor was more sensitive? Yes, I'd like to expose for 800ISO.

I still don't know what problem you have. Shoot at 400 on the cinema camera, Meter for it, use the zebras to see when it clips. It's really no different than what you or Paul had described.

Are you saying you wish the native iso was less noisy?
Are you saying you wish the camera had a lower native iso?

Ok - then really this has little to do with being able to expose to the right or not. You just want a cleaner image at asa 800.

If you nail exposure at 800 - with a touch of noise reduction it can look very clean.

Example:

Raw 800 - No Noise Reduction

http://www.tommajerski.com/publicimages/800_noNR.jpg

Raw 800 - Moderate Noise Reduction

http://www.tommajerski.com/publicimages/800_NR.jpg

Yes, given my experience so far with the BMCC, I wish it were less noisy, but as pointed out I may be able to do better in this regard.

Do I wish the sensor was more sensitive? Yes, I'd like to expose for 800ISO.

I still don't know what problem you have. Shoot at 400 on the cinema camera, Meter for it, use the zebras to see when it clips. It's really no different than what you or Paul had described.

Are you saying you wish the native iso was less noisy?
Are you saying you wish the camera had a lower native iso?

Ok - then really this has little to do with being able to expose to the right or not. You just want a cleaner image at asa 800.

If you nail exposure at 800 - with a touch of noise reduction it can look very clean.

Example:

Raw 800 - No Noise Reduction

http://www.tommajerski.com/publicimages/800_noNR.jpg

Raw 800 - Moderate Noise Reduction

http://www.tommajerski.com/publicimages/800_NR.jpg

I like the pre-noise reduction as it looks less 'plastic' and more natural.....am thinking at some stage we will reach an era called 'post-detailism'.

yea I prefer the pre-nr too - remember that this is also still a 2.5k image - after down scaling to HD and export compression - a lot of the noise is hidden far more naturally too.

yea I prefer the pre-nr too - remember that this is also still a 2.5k image - after down scaling to HD and export compression - a lot of the noise is hidden far more naturally too.

Yeah, that is exactly what I was thinking....the grain created when shifting to HD is really worth every cent spent on this camera. Noise reduction is useful for Sky's, Blacks and shiny cars...apart from that, everything else looks way more natural as part of the shift from high res to HD, as the grain that is created is way too cool for school.

What I especially like about this grain is that it gives the image texture, which is like being able to feel the difference between photo copy paper and archival hand torn art paper....the hand torn archival art paper feels way better to touch.

One day, before I die...am going to make a movie with hand torn archival art paper (;->>

I understand, I think, this ETTR practice.

In fact, I did this with my B&W still film... I would shoot most Tri-x, HP-4, Neopan 400, at ASA/ISO 200. Since I had control over the processing, doing it all my self, I would compensate for the 'over exposure' but cutting development time. How much cut was determined by negative densitometer readings.

That didn't spare me burning and dodging in the printing phase... but it did allow for much better shadow detail, while still retaining highlight detail.

I have done some tests with my BMPCC and have 'overexposed' by two stops, and brought the image back in Resolve, with good results.

But even when I've under exposed the BMPCC, I've not been put off by the 'shadow noise'... there are cameras I've used, that produce shadows that look like they are in a black snow storm... I think for the BMPCC with a 'film grain' filter, it would mask any digitalness for most viewers... pixel peepers will always complain...

I don't know if motion picture photographers did this sort of 1/2 the rated ISO value, or not. But I do know that in 1959-60 there was a grand rerating of all film speeds which basically doubled the ASA rating.

So, if someone used the 'old ASA', they would be performing ETTR relative to the new ASA rating.

John, I often overexposed Plus-X or Tri-X in the 60s when I did all my own processing. Rather than cut development time, I increased my printing exposure in the enlarger. Amazing what you could bring out while dodging the highlights.


Sent from my iPad using Tapatalk

I think we're confusing "expose to the right" (ETTR) with "pushing a stop."

ETTR is the procedure of using the zebras as an exposure guide. You set the camera zebras at 100%, then adjust the aperture until you see no clipping from the zebras. You're exposing as brightly as possible.

"Pushing a stop" is to treat the camera as though it were rated one stop down, so you compensate with brighter light. Meaning the native ISO is 800, if you expose for ISO 400, you're pushing one stop.

The problem with ETTR in a narrative context is that a master and a closeup of the same talent in the same setting may have very different brightness. Exposing both as bright as possible can make for very inconsistent shots. Matching them in post can be very difficult, with poor results.

I've found for narrative projects, using a meter to determine the proper 18% grey/zone V exposure makes for more consistent shots. Rating the camera for 400 ISO (pushing a stop) often gets cleaner results.

I think someone suggested you generally expose for the scene, not the shot. In reality, these are all guidelines and not to be followed rigorously. It's art, not science.


Sent from my iPad using Tapatalk

I think someone suggested you generally expose for the scene, not the shot.

Except on nearly all narrative projects, you relight from shot to shot. Gotta get those beauty lights into the CU!

The problem with ETTR in a narrative context is that a master and a closeup of the same talent in the same setting may have very different brightness. Exposing both as bright as possible can make for very inconsistent shots.

Actually you should be able to prevent this with you lighting (and false color).

Yeah, I can appreciate film grain - texture in the out of focus part of the image. I just think it's preferable to add it in post with Film Convert and aquire clean

yea I prefer the pre-nr too - remember that this is also still a 2.5k image - after down scaling to HD and export compression - a lot of the noise is hidden far more naturally too.

Yeah, that is exactly what I was thinking....the grain created when shifting to HD is really worth every cent spent on this camera. Noise reduction is useful for Sky's, Blacks and shiny cars...apart from that, everything else looks way more natural as part of the shift from high res to HD, as the grain that is created is way too cool for school.

What I especially like about this grain is that it gives the image texture, which is like being able to feel the difference between photo copy paper and archival hand torn art paper....the hand torn archival art paper feels way better to touch.

One day, before I die...am going to make a movie with hand torn archival art paper (;->>

I find that if you expose "correctly" at ASA400, you don't need to push it as high as possible before clipping. You can just expose so your image looks right.

As far as noise, I really don't see much of it at all even one stop underexposed. I never got good results pushing the exposure super high and bringing it back down in post.

I take this perspective in regards to ASA and noise:

On a Canon 7D I would never shoot at or above 1250 ISO. 640 was as high as I would go and I didn't like going that high. I mainly stuck to 320 and below unless I absolutely had to, but never 1250. The noise was too unacceptable. For me, the BMCC's 1600 is the new 1250. Sure the 7D can go to 3200 or 6400 but I'd never ever shoot that high because of the noise.

I'm going to take your advice Paul (and others) and experiment next week.

I find that if you expose "correctly" at ASA400, you don't need to push it as high as possible before clipping. You can just expose so your image looks right.

As far as noise, I really don't see much of it at all even one stop underexposed. I never got good results pushing the exposure super high and bringing it back down in post.

I take this perspective in regards to ASA and noise:

On a Canon 7D I would never shoot at or above 1250 ISO. 640 was as high as I would go and I didn't like going that high. I mainly stuck to 320 and below unless I absolutely had to, but never 1250. The noise was too unacceptable. For me, the BMCC's 1600 is the new 1250. Sure the 7D can go to 3200 or 6400 but I'd never ever shoot that high because of the noise.

I've never used false colour. I've avoided any aid that affects my view of the frame too much. For example, I know many of my colleagues who do the same job as me use quite a lot of peaking to judge focus (it's TV). I try and get away with as little as I can cope with. Also doing ENG, or Portable Single Camera where I came from, most TV ops expose with the help of zebras. Although I've not done much professionally, when I have I set zebras to 85% and 'chase' them off the (Caucasian) skin leaving little trace of zebras in the view finder. I remove all battery, media, audio etc indicators. My point beng, I like to have the viewfinder clean of anything, so I can eye the proportions of composition unhindered. Sorry I'm going on...so does false colour display all the time or just display when there's a problem?

The problem with ETTR in a narrative context is that a master and a closeup of the same talent in the same setting may have very different brightness. Exposing both as bright as possible can make for very inconsistent shots.

Actually you should be able to prevent this with you lighting (and false color).

I've never used false colour. I've avoided any aid that affects my view of the frame too much. For example, I know many of my colleagues who do the same job as me use quite a lot of peaking to judge focus (it's TV). I try and get away with as little as I can cope with. Also doing ENG, most TV ops expose with the help of zebras. Although I've not done much professionally, when I have I set zebras to 85% and 'chase' them off the (Caucasian) skin leaving little trace of zebras in the view finder. I remove all battery, media, audio etc indicators. My point beng, I like to have the viewfinder clean of anything, so I can eye the proportions of composition unhindered. Sorry I'm going on...so does false colour display all the time or just display when there's a problem?

The problem with ETTR in a narrative context is that a master and a closeup of the same talent in the same setting may have very different brightness. Exposing both as bright as possible can make for very inconsistent shots.

Actually you should be able to prevent this with you lighting (and false color).

False colour wont affect you composing the shot - its an overlay onto the image - rather than a gui meter which takes up space.

There are different flavours of false colour - some which show steps by say every 10-20% of exposure, Others which show 0-16 50% and then 90%+ etc. False colour is - for my money - the best exposure tool. It gives you a massive amount of important data without blocking the view - it gives you a complete overview of the scene in real time. Waveform is second best in my book - but its spacial data is only in 1 direction. False colour makes lighting ratios very very easy to nail.





You would toggle it on or off to check your lighting - so it need not be on the entire time.

Yeah that's brilliant Tom. So you can see where the fill side of the face is exposed compared to the keyed side. I could use it at work to study the ratios lighting directors I'm working with are using as well as getting to grips with it for my own projects. And as you say, toggling it on and off would keep the EVF unclutered. Methods I've used till now have only indicated the exposure of the brightest region leaving fill as a guess/reliance on a calibrated monitor/EVF. Thanks

Do manufacturers use different colours? Here's a key from my Alphatron EVF manual. It's a shame if it becomes second nature learning the colours for one system and then update your hardware and have to learn again.

Do manufacturers use different colours? Here's a key from my Alphatron EVF manual. It's a shame if it becomes second nature learning the colours for one system and then update your hardware and have to learn again.

Yeah they do use different colors. Some will have options to match other manufacturers but many of them use their own colors.

Do manufacturers use different colours? Here's a key from my Alphatron EVF manual. It's a shame if it becomes second nature learning the colours for one system and then update your hardware and have to learn again.

From what I saw at NAB, the BMD EVF was using the Arri color scheme. A few manufacturers use this or something very close to it like Arri (clearly), SmallHD, Odyssey, Marshall, etc.