Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

RED S35 or Panasonic AF100

Status
Not open for further replies.
It seems that one of the major attributes for RED users is the REDRAW codec. Is there any sticky thread where I (as a RED newbie) can read and learn about the benefits of such a codec in comparison with others?
 
:beatdeadhorse5:


Arguing over two camera that are unavailable is a little pointless don't you think?

You might as well start a heated discussion over which super hero has the best powers for crying out loud!
 
Nelson Alexandre said:
It seems that one of the major attributes for RED users is the REDRAW codec. Is there any sticky thread where I (as a RED newbie) can read and learn about the benefits of such a codec in comparison with others?
Click to expand...

The best way to learn about Redraw is to download a copy of the latest Redcine X software and find a few .r3d clips to play with. You will learn more that way than reading anything.
There are some samples available in the work flow section at Scarletuser.com
 
David W. Jones said:
:beatdeadhorse5:


Arguing over two camera that are unavailable is a little pointless don't you think?

You might as well start a heated discussion over which super hero has the best powers for crying out loud!
Click to expand...

It is also pointless to compare a 13mp full resolution raw motion camera with a high quality intra frame unencoded recording system with a 2.2mp HD camera shooting a long GOP consumer delivery codec.
 
David Rasberry said:
The best way to learn about Redraw is to download a copy of the latest Redcine X software and find a few .r3d clips to play with. You will learn more that way than reading anything.
There are some samples available in the work flow section at Scarletuser.com
Click to expand...

That's why I love forums :) Cool tip. I thank you a lot.
 
As long as we are on the topics of cameras that could/might exist I wish someone would release a camera with a similar form of oversampling and down-resing on a single chip as the Panavision Genisis. I feel the single sensor bayer chips bloat the filesize of images. A 1080p 4:4:4 image created from one of these non-bayer chip sensors would appear sharper, because of the increases color information, then a 4K Bayer (even in raw) that was downresed to 1080p 4:4:4, the Bayer, of course, would look sharper when both where displayed at 4K.

A Genisis style senor would also have the benifit of being able to output a much smaller and easier to work with file, that is around the same quality as 4K Red Raw.

Think of it as Foveon vs Bayer, except oversampling to make it like a Foveon instead of stacking three sensors on top of eachother.
 
Chris Gold said:
As long as we are on the topics of cameras that could/might exist I wish someone would release a camera with a similar form of oversampling and down-resing on a single chip as the Panavision Genisis. I feel the single sensor bayer chips bloat the filesize of images. A 1080p 4:4:4 image created from one of these non-bayer chip sensors would appear sharper, because of the increases color information, then a 4K Bayer (even in raw) that was downresed to 1080p 4:4:4, the Bayer, of course, would look sharper when both where displayed at 4K.

A Genisis style senor would also have the benifit of being able to output a much smaller and easier to work with file, that is around the same quality as 4K Red Raw.

Think of it as Foveon vs Bayer, except oversampling to make it like a Foveon instead of stacking three sensors on top of eachother.
Click to expand...

If I'm not mistaken, the perceived sharpness of the Genesis/F35 sensor is due to aliasing rather than true resolving power. A quality offline downscaler working with a 4k bayer image will always produce an image that resolves more detail than an on camera solution.

Second, file size is the result of many factors, and you must remember that a bayer image is only one channel, whereas an RGB or similar image is multi-channel, hence multiplying the file size for each additional channel. Beyond that, there is also the issue of compression...I believe I've read that 4k REDRAW is about the same bitrate as SD uncompressed video.

And finally...as it turns out Epic/Scarlet will have the option to record to 1080p RGB downsampled from the whole sensor, if that's what you really need. :)
 
Gabe Conroy said:
If I'm not mistaken, the perceived sharpness of the Genesis/F35 sensor is due to aliasing rather than true resolving power. A quality offline downscaler working with a 4k bayer image will always produce an image that resolves more detail than an on camera solution.

Second, file size is the result of many factors, and you must remember that a bayer image is only one channel, whereas an RGB or similar image is multi-channel, hence multiplying the file size for each additional channel. Beyond that, there is also the issue of compression...I believe I've read that 4k REDRAW is about the same bitrate as SD uncompressed video.

And finally...as it turns out Epic/Scarlet will have the option to record to 1080p RGB downsampled from the whole sensor, if that's what you really need. :)
Click to expand...

Both types of sensors need to de-alias, though the Genesis needs to do it less so because of its sensor design.

It has the same amount of green, red, and blue photosites which means that with just 3 times the amount of pixels of HD it will have RGB info for each pixel. This amounts to around a "4K" sensor.

A bayer sensor typically dedicates around twice as much of its photosites to green/luma as it does to red and blue, this means with a 4K sensor there is more Green information then the a Genisis type 4k sensor, but less blue and red information. The human eye is much better at seeing green so it is less noticable most of the time, but sometimes (in solid red obects especially) the lack of color resolution can be very noticable. And when a 4k Bayer is downsampled to 1080p it loses its sharpness advatange in the green, and still doesn't have R & B for every pixel, so it is at a disadvantage.

To get the same amount of R & B color information as a 4K Genesis type (I say Genesis type because it is the only camera I know of that uses this arrangement of photosites) you need a 6K Bayer, but if all you are doing is downresing to 1080p then you are having to deal with larger file sizes and decreased light sensitivity and dynamic range then if the sensor was not a bayer type.

For cameras like the Red One and Epic it makes sense to shoot with a bayer, because they are more likely to be projected in higher then HD resolution, where the bayer will look much sharper.

But for the under 10,000 35mm video cameras, it would make more sense to arrange the rgb of the photosites to output a very strong 1080p, rather then a 4k bayer that will look (if only a little) worse once downsampled to 1080p.
 
Ok. So I became aware of a new detail reading another forum, and this is the kind of info that being true puts an end to my question, which is RED or Panasonic...

I'm pasting the other thread here (I hope that I'm not prevaricating on any rule here doing so). Can anybody confirm if the following is true?


"So, we start with the most affordable Scarlet fixed lens $4750 with redcode 42.

First option: 128 GB SSD, formatted space is probably about 119GB which is 47minutes at 42MB/sec.

If 47minutes is enough, that's $1500 for SSD. Total cost $6,450 with SSD reader.

If you need 94 minutes, that $3,000 for SSD. Total cost is $7,950 with SSD reader.

the $300 CF is even more expensive for these capacities.

So, the cheapest Scarlet is now a $8,000 camera and that's without any addons.



Now lets see an EPIC FF35:

It records at 225MB/sec.

Assuming the $1,500 1.8" SSD can do 225MB/sec write, it can hold about 9minutes of footage.

The 90 minutes will cost $15,000."
 
Nelson Alexandre said:
Ok. So I became aware of a new detail reading another forum, and this is the kind of info that being true puts an end to my question, which is RED or Panasonic...
Click to expand...

The actual data rate varies with frame rate and compression settings. The published approximations are the maximum rates one would encounter at the highest quality and frame rates. By the way Scarlet's max has been upped to Redcode 100.

Yes it is going to take a lot more storage space to record any where from 2-1/2 times to 6 times the pixel data per frame than would be the case with HD, especially highly compressed long GOP video where you are in reality only recording 1 out of about every 5 to 15 frames at full resolution and detail. A 4k Red 1 clip at 24 fps Redcode 36 I have on my laptop takes about 1.4 gigabytes for less than a minute. Such is the price of true uncompromised original image quality.
 
You can get 64GB CF cards for $200 or less. That is 24 minutes in Redcode 42. Besides, seeing as how for 4K on the first gen camera the difference between Redcode 42 and Redcode 28 is minimal, I don't see why we would be forced to shoot Redcode 42 for 3K Scarlet footage.

Same with the Epic. Why should you be forced to shoot 225? That is approaching Redcode lossless, isn't it? Much cheaper than Alexa's $60k RAW recorder plus the media, I would say.
 
David Rasberry said:
The actual data rate varies with frame rate and compression settings. The published approximations are the maximum rates one would encounter at the highest quality and frame rates. By the way Scarlet's max has been upped to Redcode 100.

Yes it is going to take a lot more storage space to record any where from 2-1/2 times to 6 times the pixel data per frame than would be the case with HD, especially highly compressed long GOP video where you are in reality only recording 1 out of about every 5 to 15 frames at full resolution and detail. A 4k Red 1 clip at 24 fps Redcode 36 I have on my laptop takes about 1.4 gigabytes for less than a minute. Such is the price of true uncompromised original image quality.
Click to expand...

Subhadip Sen said:
You can get 64GB CF cards for $200 or less. That is 24 minutes in Redcode 42. Besides, seeing as how for 4K on the first gen camera the difference between Redcode 42 and Redcode 28 is minimal, I don't see why we would be forced to shoot Redcode 42 for 3K Scarlet footage.

Same with the Epic. Why should you be forced to shoot 225? That is approaching Redcode lossless, isn't it? Much cheaper than Alexa's $60k RAW recorder plus the media, I would say.
Click to expand...

David and Subhadip, one thing got me thinking:

If RED compression scheme is so flexible and efficient, why not tone it down for storage purposes?

I could easily see a REDCODE 10 working for 3K footage if you wanted to use the least space and bandwidth possible.

So people could use cheap CF cards and only use high-grade ones for high speed.

IMO that would be a smart move for the Scarlet market.
 
I think I remember hearing that the redcode numbers like 100 and such for the new line had no direct correlation to the MB/sec anymore, but I could be wrong. Either way all that assuming of how much space 3k footage will take up is just that assumptions IMO we should just wait until the full stats are out from the red team before jumping to any conclusions.
 
Indeed, Pietro, it would be a good idea for the operator to dial in the bitrate and perhaps even bit depth. I remember Jim shot some 5K 2.35:1 120fps footage in REDCODE 10 on an Epic and mentioned how good the image looked. I also remember him mentioning that REDCODE 2 is better than the Canon DSLR H.264 codec. (I am just speaking from memory, so don't quote me)

5K 120fps is at least 11 times as much data compressed, compared to 3K 24fps. Perhaps even more if the Scarlet does not shoot 16-bit like the Epic. There's no reason why 3K 24 fps footage wouldn't look good at REDCODE 10, then. I am sure they will offer many more REDCODE options for Scarlet, though I don't know the precision. But, sure, REDCODE 10 could be possibility. By the way, the REDCODE nomenclature will be changed to ratios before Epic's release.

REDCODE 100 is just an option - and is likely to be overkill. You aren't forced to use it. Neither are you forced to use 42. You can get by on a CF card costing a few hundred dollars just fine. The concerns in the previously quoted post are unfounded.
 
Jim Retar said:
Will they work on RED though?

I would expect the 64gb RED card to be $800+
Click to expand...

It is likely they won't work at Redcode 42, you are right. Like I have been saying, you are not forced to shoot at such a high bitrate, though.

Some $200 64 GB CF card are 533x, which is rated at about 70 MB/s, I believe. Realistically, you can expect a minimum write speed dips of about 30-40 MB/s, or so. Redcode 28 could be doable on these cards, no? A top of the line 64GB 600x card is around $600 - and these are verified performing with dips of above 55 MB/s. On these, even Redcode 42 should be no issue.

Of course, it wouldn't be as safe as RED approved media, but you aren't forced to use that either. All I am saying is there are plenty of options. Spending $8,000 on media is really not obligatory.
 
There are issues with some CF card memory controllers that make them incompatible with the way Red writes files, so fast or not many won't work. Maybe the new cameras will be more flexible in this respect. If it is like Red 1, the camera will test the card during format and limit the camera settings to a safe data rate for the card.
There is a reason most camera manufacturers build their own certified media to insure adequate continuous write speeds and reliability under repeated use.
 
The Panasonic AF-100 will do 1080 24p albeit the much "low end" format (no point in comparing here, too different).
The question is that the AF-100 will have a 4/3 sensor (quite alright) and it will be available 'way' before the the Scarlet S35.
I predict a lot of people will be jumping on that wagon.
 
As OP I've been reading through all this and other posts to make an educated choice (expecting both cameras to be released at the end of this year).

In my quest I've stumbled across a bunch of info I was not aware (thank you forums), as RED using T-stops for instance. How's this important for a choice between this 2 cameras you might ask?

Besides a bunch of Nikon primes (starting at my fastest 50mm 1:1.2), I was expecting to use zooms more regularly. I have 2 Sigma wide zooms I was planning to use with the AF-100 which are 10-20's f/3.5 and 24-60's f/2.8.

The 24-60's 2.8 aperture is fine but it translates into a normal-to-mid tele lens with a 4/3's sensor, which cannot be used in tight spaces. The 10-20 will work, but its aperture will probably translates into a real T/4 lens, so rapidly approaching the shallow DOF permitted by Scarlet's 2/3 sensor with the announced T/2.8 lens.

You can be thinking why I don't use a fast wide prime (which I have), in those tight spaces I would probably be shooting at, and the reason is that I mostly do documentary work, direct cinema style, so hence my need for the zoom lens instead.

Other issue is... What Scarlet? Brain only or Fixed lens? If brain only, I don't have any zoom lens which will work as a wide lens on its 2/3 sensor.

On a side note, Scarlet size will be great for my kind of work, but it really seems a toy compared to a AF100 (in the worst kind of way). I just don't know why RED doesn't build their own cameras to look like a camera, instead of a futuristic water pistol... (no harms done here, just a though with my excuses in advance for all of you RED fans).
 
Status
Not open for further replies.
Back
Top