BMPCC4K - BRAW Best Format

Definitely know my bits. Software dev.. but thank you for pointing that out. I didnt mean that it was just 2bits more of data like if it was only 2 bits. I realize that 2bits adds a lot more color/etc to the equation. I know 12bit is a lot better than 10bit, which is vastly better than 8bit. Also the 444/ 442/ 420, etc makes a difference.

I dont recall though if BRAW is 12bit 444 or 422 or what?

I dont recall though if BRAW is 12bit 444 or 422 or what?

To my knowledge, it's 12bit raw, so color subsampling shouldn't apply.

Definitely know my bits. Software dev.. but thank you for pointing that out. I didnt mean that it was just 2bits more of data like if it was only 2 bits. I realize that 2bits adds a lot more color/etc to the equation. I know 12bit is a lot better than 10bit, which is vastly better than 8bit. Also the 444/ 442/ 420, etc makes a difference.

I dont recall though if BRAW is 12bit 444 or 422 or what?

It meant to be 12bit but it's most likely not pure RAW Bayer data, but half pre-processed. In terms of color sampling then RAW is not 4:4:4, nor 4:2:2- it's Bayer data. After debayering it gives you 4:4:4 data, but not perfect one as some info is interpolated (if your oversample then you'll get close to 4:4:4). BM RAW is the same. You can think about it as 4:4:4 with some info being interpolated.

Will I need the CFAST for BRAW? Or can I use lower cost compact flash memory cards?
.

CFAST and Compact Flash are not interchangeable. Despite being the same form factor, they have very different pins. The BMCC4K does not have a memory slot that's compatible with Compact Flash cards.

How much faster is rendering and working with BRAW than CDNG?

To my knowledge, it's 12bit raw, so color subsampling shouldn't apply.

Also import to note is that it’s log encoded 12bit raw data, not simply linear 12bit raw data

ARRI Alexa records 12bit log encoded raw as well, and according the color scientists at ARRI, 12bit log data is functionally equivalent to 16bit linear data

https://www.arri.com/en/learn-help/learn-help-camera-system/camera-workflow/image-science/frequently-asked-questions-on-hdr#accordion-44150

I would say perceptually equivalent. Technically, it’s still throwing some data away.

As with most things, there’s no blanket rule that will cover every situation.

I recently shot interviews at 12:1 and the quality was great. I didn’t have unlimited drive space for this and needed to shoot well over an hour. I’d say 8:1 and 12:1 look about the same.

If you are shooting a giant landscape with tons of detail, you will see a difference between. 12;1 and Q0. For situations like that, I’d do Q0 or a higher constant nitrate. But it will require testing for people to make their own decisions. I know plenty of shooters that love 12:1 for people/faces no matter what. I don’t think I know anyone shooting Q0, but it’s there for people who need it.

cDNG is better. Bring it back.

Can we revisit this topic now that we've had time to shoot? What does everyone think? I normally shoot with 5:1, but I'm considering going to 3:1.

Can we revisit this topic now that we've have time to shoot? What does everyone think? I normally shoot with 5:1, but I'm considering going to 3:1.

I use 8:1 for everything on the pocket 4K now, except 120P I will do 3:1 braw, but I hardly ever shoot 120p. Even with fine hair detail on insects 8:1 looks great in my final renders.

Also use 8:1 here although I wish there was a Q3 available.

I find with 5:1 and slight sharpening, the image is very CDNG-ish. I think a 0.47 sharpen, not in front of my workstation so I don't remember the exact name of the setting, but I'm talking Davinci Resolve. I'll probably stick with 5:1, but I'll give 8:1 a shot again.

Also, is anyone using Constant Quality?

.48 sharpen with 8:1 gives you a very sharp image with less haloing than the .47. FOr me, .48 is the sweet spot. If you downscale to HD, you won't need any NR and it looks really, really good.

I find with 5:1 and slight sharpening, the image is very CDNG-ish. I think a 0.47 sharpen, not in front of my workstation so I don't remember the exact name of the setting, but I'm talking Davinci Resolve. I'll probably stick with 5:1, but I'll give 8:1 a shot again.

Also, is anyone using Constant Quality?

.48 sharpen with 8:1 gives you a very sharp image with less haloing than the .47. FOr me, .48 is the sweet spot. If you downscale to HD, you won't need any NR and it looks really, really good.

I find with 5:1 and slight sharpening, the image is very CDNG-ish. I think a 0.47 sharpen, not in front of my workstation so I don't remember the exact name of the setting, but I'm talking Davinci Resolve. I'll probably stick with 5:1, but I'll give 8:1 a shot again.

Also, is anyone using Constant Quality?

I agree that you don't need much, but I stack diffusion filters. So, I typically like .47, it doesn't look too sharp when using the diffusion filters.

I’m usually shooting Q0 or ProRes 422 HQ. Compare that to CinemaDNG 3.1 and ProRes 444 unless I’m shooting for extended hours.


Sent from my iPhone using Tapatalk

I am with Ulli though unlike him I am not able to contribute to the technical issues.

I am a long time Red User (its funny now seeing people from that forum now on the Blackmagic site in increasing numbers) and I remember similar kinds of debates in the early days of Red with some people questioning the "softness" of Raw. There were those who craved detail and sharpness and there were others who just hated that look. I have used my Epic X now for years and it has a very pleasing and what I would describe as kind of organic softness that makes the images look very natural. I hardly ever add any sharpening of any kind in post.

I am looking forward to seeing how this debate unfolds over time.

Rob

Can we revisit this topic now that we've had time to shoot? What does everyone think? I normally shoot with 5:1, but I'm considering going to 3:1.

Any reason to not do that?

For instance, I have bought x2 1TB Samsung T5 and this gives you me an option to edit directly from the drive and have another for a backup. I have spend ~300$ (tax included) on it. For professional use - space is not an issue anymore, unless you want to keep an archive of BRAW.

I am shooting Q0 and it gives me an option to record 120 minutes 4k 24 on 1 TB. Unless I go to travel somewhere for a month or more, I do not see a reason to go below Q0 to save space.

IMHO

Any reason to not do that?

For instance, I have bought x2 1TB Samsung T5 and this gives you me an option to edit directly from the drive and have another for a backup. I have spend ~300$ (tax included) on it. For professional use - space is not an issue anymore, unless you want to keep an archive of BRAW.

I am shooting Q0 and it gives me an option to record 120 minutes 4k 24 on 1 TB. Unless I go to travel somewhere for a month or more, I do not see a reason to go below Q0 to save space.

IMHO

I mainly shoot music videos and I shoot a fair amount at 60fps. A typical shoot at 5:1 is around 300GB, I shot something at ProRes HQ and literally around 30 minutes was 300GB (it was all at 60fps). Storage is becoming a problem, I like to keep my files at least until the end of the year so I can do a new reel. I have 2 4TB backup drives that I offload my footage to once I'm done editing. I don't edit from my T5 (500GB), I copy the files to a interal SSD (1TB) and work from there. I plan to add another M.2 SSD (1TB) and possible work from that drive in the future. The one I have now is used as the OS drive.

So... I was asking about 3:1 because I wonder if I'll be entering the land of diminishing returns going from 5:1 to 3:1, I've done some tests and do see an increase in detail, but coming from a GH5 the 5:1 footage looks amazing. I just don't think my clients will notice a difference. One might say well just shoot in 12:1, I feel that I notice a difference at 12:1 and in my testing I ended up at 5:1. I just wanted to see if anyone was passionate about one or the other. I have never used the Constant Quality setting either. In YouTube comparisons I just can't make out a difference so I think I'll stick with 5:1. Unfortunately for me most of my work is viewed on a phone or laptop screen at the end of the day so this is really just for me.

Que, since you’re inclined to test, take Q0 and Q5 for a test drive. Jamie is very pleased with Q5 and I’ve been surprised the Q0 can consume less resources than 5:1 subject to the scene. On the flip side Q0 can consume more resources than 3:1. Takes getting used to as Q0 varies with a scene more dramatically than BMD has officially published.


Sent from my iPhone using Tapatalk

I record all I can in Q0, I had 2x4tb only to backup on set, I had a nas with 8x8tb to archive, and at today I near to fill 3/4 of space.
Today 12tb hdd are near right price.
One at month I buy new hdd and add to nas, or today I replace in nas.

Another option is a splitting of finalized work with archived used footage only, and an archive of interesting shooting for demo reel.
But is a work in a work, if you have time to do it, could be useful to clean up your storage, and reduce space.
Resolve allow you to build a new timeline, cut only good part of all shooting, and create a new original trimmed raw without re compression.
It’s possibile with braw. That reduce a space be cause if you have 2 minutes of shooting but half is out of focus, you cut out unusable shooting.


Inviato dal mio iPhone utilizzando Tapatalk

Except for VFX shots and detailed wide shots where a push in post is likely, I've been shooting everything in Q5 with zero problem. It's a dream format. If you pixel peep it can you find differences with Q0? Sure you can. The real question is, for the majority of footage that you shoot, will those differences ever matter? For the overwhelming majority of the footage that I shoot, the difference does not matter at all. BRAW Q5 is the raw format I've been waiting for and BMD delivered. By all means run your own tests, but rather than looking for the seams in contrived pixel peeping tests, just try shooting a single project with it and running it through a normal workflow to your usual distribution medium (which is most likely one streaming site or another....) and see what happens.

Shot 17 client films in Q5 since Braw arrived, and we tested the different flavors pretty extensively.

Love using it from acquisition to post...favorite codec I've ever worked with, hands down.

Very strong and knowledgeable endorsements of Q5 from Jamie and Jim. Like Carlo, I’ve been shooting Q0 when I can manage it. But I’m also reaching a crisis point with my Pegasus2 RAID 10 R6 and must manage my storage space better. I’ve got to archive old projects to offline storage. Think that will help.


Sent from my iPhone using Tapatalk

That's one of the reasons I really like Q5 Rick - efficiency and speed without any practical compromise in IQ.

So much of what we produce is with me behind the camera and I always grade and edit what what I shoot. It's amazing to me that you can shoot 4k Braw on $35 SD cards and get 40+ minutes of footage. Likewise with editing. A few 1 TB T5 drives.

For the way I work at least, Braw/Q5 has been a bit transformational.

That's one of the reasons I really like Q5 Rick - efficiency and speed without any practical compromise in IQ.

So much of what we produce is with me behind the camera and I always grade and edit what what I shoot. It's amazing to me that you can shoot 4k Braw on $35 SD cards and get 40+ minutes of footage. Likewise with editing. A few 1 TB T5 drives.

For the way I work at least, Braw/Q5 has been a bit transformational.

You guys have persuaded me to try Q5 for my next project!

In spite of using Q0 on smaller projects, you can be sure when I’m recording the musical festival here next year, I’ll be shooting Q5 as it gives me about 11 hours on a terabyte card. That’s exactly what I’ll need for the feature day, shooting continuously from 10 am to 11 pm with 10-15 minutes between acts.

If BMD brings BRAW to the URSA Mini 4.6K in the next half year or so, I have a terabyte CFast 2 card for the Mini so I can make it a dual camera shoot with the MixPre doing three or four channel recording on the SDXC 128GB card for 11 hours as well. Plus there I can use the festival’s professional sound board audio recorded to another SD card. Can’t wait.


Sent from my iPhone using Tapatalk

Except for VFX shots and detailed wide shots where a push in post is likely, I've been shooting everything in Q5 with zero problem. It's a dream format. If you pixel peep it can you find differences with Q0? Sure you can. The real question is, for the majority of footage that you shoot, will those differences ever matter? For the overwhelming majority of the footage that I shoot, the difference does not matter at all. BRAW Q5 is the raw format I've been waiting for and BMD delivered. By all means run your own tests, but rather than looking for the seams in contrived pixel peeping tests, just try shooting a single project with it and running it through a normal workflow to your usual distribution medium (which is most likely one streaming site or another....) and see what happens.

I'm going to shoot a project coming up on my G2 using Q5 (I've only used 5:1 so far of the BRAW formats). It's possible I shoot some slow motion. Do you still recommend the Q5 for the 120 fps or does the faster frame rate do better with a higher bitrate?

Do you still recommend the Q5 for the 120 fps or does the faster frame rate do better with a higher bitrate?

I haven't seen any difference in Q5 performance between 24fps and 120fps. Same rules apply regardless of the frame rate.

thanks!

Resolve allow you to build a new timeline, cut only good part of all shooting, and create a new original trimmed raw without re compression.

Hey Carlo, could you please share how to achieve that in resolve? I haven't seen an option to export in BRAW neither delivery page nor project archive. Thanks in advance!

Resolve allow you to build a new timeline, cut only good part of all shooting, and create a new original trimmed raw without re compression.

Hey Carlo, could you please share how to achieve that in resolve? I haven't seen an option to export in BRAW neither delivery page nor project archive. Thanks in advance!

It's under File / Media Management. Chapter 37 of the manual.

Is Q5 better than 12:1? I been sticking to 12:1 for my sports video stuff. Seems to take up about 250GB or so for a bit over 1.2 hours of video. Should I use Q5 instead? Mostly shoot 30fps 4K DCI.

You won’t know until you try but for sports, Q5 might actually consume fewer resources than 12:1. It ranges as low as about 20:1 while keeping the detail you likely need. Just try it in a very similar situation and see the results.


Sent from my iPhone using Tapatalk

Another vote for Q5 here.

Did my own tests like many others did. It's really the only thing that's worth anything, because people might have wildly different expectations.

In pixel peeping tests I'm scratching my head because it holds up so well. Every now and then I look at the technical specs of it (compression) and get new doubts and do some more peeping. But then I shoot some more and re-confirm what I already know.

So, for any project where you want to have the camera rolling (like documentary-ish), it's really great and a no brainer for me.

If I was planning to do some continuous light photography and grab frames.... well, then I'd just go Q0 because of the small scope of the project—it's still good to have options.

I'm pretty sure Braw Q5 is the overall best codec out there—counting everything there is—in terms of quality, flexibility and efficiency (when all are equally important).

Rick, that is interesting. I thought Q5, due to the nature of sports having a lot of motion, would end up with more data, larger files, etc at.. in my case little to no benefit in terms of overall quality. I'll have to try Q5 now, I been using 12:1 all this time as it is way way better than the MP4 I record with handycam (when I am stuck to using that).

BRAW is an I-frame only codec. Motion doesn't really matter (apart from some blur).

Alright, after the full press efforts to promote BRAW Q5, I actually did use Q5 to record a musical theatre production on Sunday. Now I must say, I had the CFast 2 capacity to record Q0, but I guess I felt I owed it to those who voted for Q5 as the best codec.

The actual length i recorded was about two hours and forty minutes and the clips totaled under 80GB! I’m used to recording 500GB or more for less duration. I did do some tests beforehand and as much as my intellect tells me Q0 has to be better, I couldn’t conclusively see a difference between Q0 and Q5 shooting some static objects in my ‘studio.’ I’m in the process of editing the Q5 video and it seems alright.

Having Timecode synced video and dual system 32bit audio via the MixPre-6 II and getting an XLR line feed from a pre-production soundboard of a new fancy model was like manna. Don’t ever want to go back.

I did record audio from the BMPCC4K internal mics and it was pretty decent. I even had it set to 50% (no gain) and it picked up a lot of sound. I recorded it in case there was an unforeseen disaster but I don’t need to use it as the XLR line feed and the Sennheiser MKH416 did a great job.

Allow me one more aside. Since I was already expecting the worse, I took another gamble. I had been thinking that the Tokina 11-20mm T2.9 would require either ISO 1250 or 3200 to get a ‘good exposure’ for BRAW. And we all know what a good exposure looks like by eyeballing the histogram. You want to see that white histogram hill spread out at least around the mid-point of the scale if not somewhat to the right of the scale.

Now some people have felt that ISO 3200 is noisy (especially in a dark theatre with most scenes dark or under lit) and so I decided at the last moment to use ISO 400. But I also needed reasonable depth of field as the stage is forty feet wide and nearly as deep. There is a very large cast using the entire area and I wanted them all in focus. So I went with 20mm to handle the stage width and T5.6 to manage the depth. But that histogram was mostly in the lower left quartile or less. That was the gamble. To go with an intentionally underexposed scene. After preliminary first light of the video today, it’s all good! It’s Blackmagic.

Rick, that is interesting. I thought Q5, due to the nature of sports having a lot of motion, would end up with more data, larger files, etc.

Consider the following:

a Long GOP format Finter Frame compression) that shoots a stationary frame from a tripod can take advantage of fact that each frame in a long sequence is more or less the same as the previous one, or the next one. This format optimises file size by taking change over time into account, so if there is little change, files become smaller.

An Intra Frame compression like Braw or ProRes, that stores individual frames, is not influenced by motion in preceding or succeeding frames. This is great for 'motion stability' when shooting moving water or foliage that moves in wind, for example. Since all compression takes place 'inside' the frame, the question becomes: how much fine detail does the frame hold that needs encoding?

Filming sports in an intra frame codec like Braw with a filmic shutter speed will give you frames with a fair amount of motion blur, which in turn will result in easily compressed frames.

Additionally for the musical theatre performance I just shot, there are many low lit shots with only a few characters on stage and Q5 can jump on that and take advantage of those areas that don’t have much detail to them whereas the actors are still captured in sufficient detail. The
recording may fall below the equivalent of 20:1.

When everyone is on stage and moving, there’s even more action in the frame than most sporting events and that’s when Q5 may effectively be recoding better than 7:1.


Sent from my iPhone using Tapatalk

Hi Guys

Really interesting stuff here, and great read!

I am sadly camped in FCP X from an editing point of view - so the lazy part of me wants to stay with Prores 422 or HQ for direct import into FCP X.

I am not adverse however to shooting in Q5 - bringing into resolve in log and then delivering straight back out again as prores hq to then import back into fcp x..

my question is - am I gaining anything from doing the latter in terms of IQ from the Q5 or am I losing any gain as soon as I convert out to prores hq?

Rick, that is interesting. I thought Q5, due to the nature of sports having a lot of motion, would end up with more data, larger files, etc at.. in my case little to no benefit in terms of overall quality. I'll have to try Q5 now, I been using 12:1 all this time as it is way way better than the MP4 I record with handycam (when I am stuck to using that).

Data-rate in QX modes will be mainly down to the amount of noise in the shots. Indirectly amount of movement in the frame also matters (as it translates to how complex is every frame) but much less. Well exposed shots with plenty light should come up clean and at lower rate then any low light ones (for the same Q mode).

Hi Guys

Really interesting stuff here, and great read!

I am sadly camped in FCP X from an editing point of view - so the lazy part of me wants to stay with Prores 422 or HQ for direct import into FCP X.

I am not adverse however to shooting in Q5 - bringing into resolve in log and then delivering straight back out again as prores hq to then import back into fcp x..

my question is - am I gaining anything from doing the latter in terms of IQ from the Q5 or am I losing any gain as soon as I convert out to prores hq?

Hello Lee.

Generally speaking it's never great to have codec concatenation, meaning, whenever you convert from one codec to another there's an inevitable loss in most situations that involve compression.

Q5 by itself is an amazing codec. ProRes is as well. I think the only way you'll know if if you test it.

You COULD also go to ProRes 444 if you want to eliminate the transcode as an issue and stay 12 bit, but that's going to generate much larger file sizes.

If you're "happy" with ProRes HQ, I' would guess that you might be OK with Q5 transcodes, but it's very much going to be dependent on your source material, how much you expect to be able to grade and alter it, and your final delivery specs....

JB

Hi Guys

Really interesting stuff here, and great read!

I am sadly camped in FCP X from an editing point of view - so the lazy part of me wants to stay with Prores 422 or HQ for direct import into FCP X.

I am not adverse however to shooting in Q5 - bringing into resolve in log and then delivering straight back out again as prores hq to then import back into fcp x..

my question is - am I gaining anything from doing the latter in terms of IQ from the Q5 or am I losing any gain as soon as I convert out to prores hq?

Hello Lee.

Generally speaking it's never great to have codec concatenation, meaning, whenever you convert from one codec to another there's an inevitable loss in most situations that involve compression.

Q5 by itself is an amazing codec. ProRes is as well. I think the only way you'll know if if you test it.

You COULD also go to ProRes 444 if you want to eliminate the transcode as an issue and stay 12 bit, but that's going to generate much larger file sizes.

If you're "happy" with ProRes HQ, I' would guess that you might be OK with Q5 transcodes, but it's very much going to be dependent on your source material, how much you expect to be able to grade and alter it, and your final delivery specs....

JB

Hi John many thanks for your response!

I think for me the issue also is storage. I am trying to go as naked and as cheap as possible right now so want to shoot straight to sd cards...shooting Q5 or even Q12 will allow me to do this - where as ProRes is a hungry beast and I probably wont be able to get that on my sd cards. if I could I would probably just shoot in prores in cam and then stay in fcp x!

Thanks!

Lee

I haven’t shot much at all with action. Depending on your shutter angle and frame rate, your frame may have a lot of motion blur and that’s not the kind of detail with edges that Q5 is going to do anything unusually good with. And it should;dn’t be bad either because the shots will often be about the motion. Definitely needs to be tested and compared, but if you know media use needs to stay low, it’s what I’d shoot. When I have the space, I shoot Q5 and ProRes 444 if it’s available.

Thanks Rick... my perfect scenario ( being a fcpx user) is to shoot in prores and edit in fcp x.. pro res eats cards for breakfast thought and even though they can handle pro res I might only get 22 mins from a 128gb card!

Therefore I am drawn to q5s data rates.. but would still need to convert that to prores hq so I could still edit in fcp x.. but I don’t want to lose any quality in the conversion!

Thanks Rick... my perfect scenario ( being a fcpx user) is to shoot in prores and edit in fcp x.. pro res eats cards for breakfast thought and even though they can handle pro res I might only get 22 mins from a 128gb card!

Therefore I am drawn to q5s data rates.. but would still need to convert that to prores hq so I could still edit in fcp x.. but I don’t want to lose any quality in the conversion!

the good of braw is that you can load easely in resolve, export the prores that you want, and add a bit of color immediately or later, but you start from a 12bit log stuff to 10bit prores instead to shoot directly in 10 bit prores.
You must do you test, also if i want to edit on a nle that not read braw, i would like to shoot in braw, more infos, more chance to adjust in post some shooting error.

If you do some rough grading (a so-called first light) to get it into the ballpark in Resolve before conversion, you should be fine.

If you do some rough grading (a so-called first light) to get it into the ballpark in Resolve before conversion, you should be fine.

Thanks for the feedback.. I normally try to shot and expose properly in camera so my log footage normally doesn’t need tweaking that much..

That being said.. in terms of adding luts.. would advise to add a lut to the footage in Resolve and then export to proreshq or export the ungraded braw file as prores and then add the lut to in fcp x instead?

Thanks Rick... my perfect scenario ( being a fcpx user) is to shoot in prores and edit in fcp x.. pro res eats cards for breakfast thought and even though they can handle pro res I might only get 22 mins from a 128gb card!

Therefore I am drawn to q5s data rates.. but would still need to convert that to prores hq so I could still edit in fcp x.. but I don’t want to lose any quality in the conversion!

I edit most everything in FCPX and since the release of BRAW I have shot virtually everything in Q5. Best workflow for that is to pull the BRAW files into Resolve and export REC709 ProResProxy. Edit in FCPX and when the edit is complete, send FCPXML back to Resolve for grading + finish there. I've been using that workflow even with regular ProResHQ log files going back to the original BMCC with every NLE simply to make the edit simpler and not have to deal with log during the creative edit. It's a solid workflow. The only exception is if I'm doing short commercials that are mostly VFX + sound design rather than editing. Those projects I simply do completely inside Resolve.

Thanks Rick... my perfect scenario ( being a fcpx user) is to shoot in prores and edit in fcp x.. pro res eats cards for breakfast thought and even though they can handle pro res I might only get 22 mins from a 128gb card!

Therefore I am drawn to q5s data rates.. but would still need to convert that to prores hq so I could still edit in fcp x.. but I don’t want to lose any quality in the conversion!

I edit most everything in FCXP and since the release of BRAW I have shot virtually everything in Q5. Best workflow for that is to pull the BRAW files into Resolve and export REC709 ProResProxy. Edit in FCPX and when the edit is complete, send FCPXML back to Resolve for grading + finish there. I've been using that workflow even with regular ProResHQ log files going back to the original BMCC with every NLE simply to make the edit simpler and not have to deal with log during the creative edit. It's a solid workflow. The only exception is if I'm doing short commercials that are mostly VFX + sound design rather than editing. Those projects I simply do completely inside Resolve.

Thanks

Just having a read about this process.. most of the articles I have read assume you are starting in fcp x and then exporting an fcpxml into resolve.. your method starts in resolve with exporting out the proxy files.. can I ask.. if you export proxy files for use in fcp x and then edit, and export the fcpxml.. how does resolve ensure it doesn’t pick up the lesser proxy files for the timeline and use the original
Source files instead?..my confusion is it seems when importing the fcpxml back
Into resolve won’t it be importing the proxy files ??