Urban landscape with RED Epic done by Peter Lik

[harlemshooter]

The print looks good from a viewing distance of 5 feet +. Just don't get closer if you're a critic like me. That said, RED DR is quite remarkable.

[eronald]

Very easy: Lik makes $$zagillions$$, so proof is there are lots who like it...

Maybe this is what we should be learning about

Edmund

[ondebanks]

nothing that complicated..  I'm pretty sure it alternates frames... if you typically shoot a 180 deg shutter, then the sensor is off half of the time between each frame, which allows the Epic to use that "down time" to capture a quicker exposure between the normal frames which is recorded to a separate track.  The only downside is it requires twice the storage.  Oddly, every time I've turned HDRx on, I haven't ended up using the hilight protection track.  I always seem to have enough DR out of the Epic with the straight exposure.

Chris, thanks for the further explanation. It was the use of "simultaneous" that threw me. What you describe with the RED is not simultaneous; it is temporally interlaced. The cadence may be really, really high, but nevertheless the frames are captured one after the other, not simultaneously.

Ray

[feppe]

Chris, thanks for the further explanation. It was the use of "simultaneous" that threw me. What you describe with the RED is not simultaneous; it is temporally interlaced. The cadence may be really, really high, but nevertheless the frames are captured one after the other, not simultaneously.

So it's shooting 2x the designated frame rate, exposing each other frame at "normal" level, and the other at +1 to +6 stops as set by the user?

[Mr. Rib]

Huh, it bothered me as well, it had to be 'interlaced' in some way. I wonder what are the results of HDR blending of a really fast-paced footage, for instance let's say you record a car pursuit or simply shoot from the car (as in the video demonstration of HDR by Red), is the displacement between blended frames an issue.

[feppe]

Huh, it bothered me as well, it had to be 'interlaced' in some way. I wonder what are the results of HDR blending of a really fast-paced footage, for instance let's say you record a car pursuit or simply shoot from the car (as in the video demonstration of HDR by Red), is the displacement between blended frames an issue.

They created frames to get to incredible frame rates in The Matrix without artifacts, and that's now a standard feature in medium- and high-end TVs and movie projectors which do it live. I'd imagine it should be possible with few artifacts in post when rendering doesn't have to be real-time.

Twixtor is one software, and the slow-mo effect is overdone in music videos - here's a fun effect with it, though.

[Lester]

Very easy: Lik makes $$zagillions$$, so proof is there are lots who like it...

That is why RED used Lik to produce that images.

[yaya]

Perhaps someone here who's also a member on the REDUSER forum can post a few questions and try to get some straightforward answers from those who created that image;

1. What method was used for capturing that panoramic image? Was the camera turned on a tripod or perhaps this was some kind of a flat-stitch?
2. How many images were stitched? What is the size (in pixels) of the stitched EPIC image?
3. Were these images shot as stills or where they pulled from a sweep/ pan video?
4. If they were pulled from a sweep/ pan video, how long did it take to find, select, edit, process and stitch them?
5. Was the IQ180 used in a similar (or close to similar) way?
6. How many frames were used for the stitched IQ180 image?
7. Was the IQ180 used in Sensor+ mode?
8. How long did it take to select, edit and process/ stitch the IQ180 image?
9. What was the image size (in pixels) of the stitched IQ180 image
10 What lenses were used on the 2 cameras, respectively and at what aperture/ shutter speed?

The only thing that will get me really excited (not about that specific image) is if such an image could be created in the same way the little Sony P&S do their sweep+pan shots, but I think we will all appreciate it if some of the questions above can be answered

Yair

[fredjeang]

Perhaps someone here who's also a member on the REDUSER forum can post a few questions and try to get some straightforward answers from those who created that image;

1. What method was used for capturing that panoramic image? Was the camera turned on a tripod or perhaps this was some kind of a flat-stitch?
2. How many images were stitched? What is the size (in pixels) of the stitched EPIC image?
3. Were these images shot as stills or where they pulled from a sweep/ pan video?
4. If they were pulled from a sweep/ pan video, how long did it take to find, select, edit, process and stitch them?
5. Was the IQ180 used in a similar (or close to similar) way?
6. How many frames were used for the stitched IQ180 image?
7. Was the IQ180 used in Sensor+ mode?
8. How long did it take to select, edit and process/ stitch the IQ180 image?
9. What was the image size (in pixels) of the stitched IQ180 image
10 What lenses were used on the 2 cameras, respectively and at what aperture/ shutter speed?

The only thing that will get me really excited (not about that specific image) is if such an image could be created in the same way the little Sony P&S do their sweep+pan shots, but I think we will all appreciate it if some of the questions above can be answered

Yair

  I knew it !

As you have a more dicrete signature than usual, I always forgot and read you post thinkink...hey, it sounds like a Phase dealer.
(saying it with respect and kindness)

Then I saw: Leaf


Seriously: they said that one of the key of this image is a PP capacity called: Alchemy, that will be released those days in RCX. We'll know more.

[yaya]

I knew it !

As you have a more dicrete signature than usual, I always forgot and read you post thinkink...hey, it sounds like a Phase dealer.
(saying it with respect and kindness)

Then I saw: Leaf


Seriously: they said that one of the key of this image is a PP capacity called: Alchemy, that will be released those days in RCX. We'll know more.

Actually I've only set up the signature a couple of days ago:-)

Perhaps that mysterious Alchemy thing is some sort of automatic stitching from frame grabs? If not and if it took 3 hrs to shoot, select, adjust, process and stitch 6 frames to come up with a 12,000 pixel image then we can all go back to making pumpkin heads for tomorrow....

[fredjeang]

Actually I've only set up the signature a couple of days ago:-)

Perhaps that mysterious Alchemy thing is some sort of automatic stitching from frame grabs? If not and if it took 3 hrs to shoot, select, adjust, process and stitch 6 frames to come up with a 12,000 pixel image then we can all go back to making pumpkin heads for tomorrow....

But I knew your link with Leaf because MR presented you in the video when they released the big backs.

I also have a huge doubt (in fact I'm sure) that this is not going to compete with a 80 MP back. Jim exagerates evidently and he sells the buzz obviously.

The crazy thing IMO, is that it's a video camera, and that can acheive a print output not ridiculous compared to the highest still cameras. That in itself is a great accomplishement. The print quality "better" as they claim
as a 80 MP Phase back, I don't buy that either.

But it also means that it can acheive without doubt a still quality that will be enough in many commercial applications.

I hope you guys that are into the brands, are pushing the MF deciders to seriously think about convergence. I don't know if it's in the agendas but hope they won't be conservative in the future.

[yaya]

But it also means that it can acheive without doubt a still quality that will be enough in many commercial applications.

Not if it takes 3 hrs to produce....and not if the same image (more or less) can be produce at a fraction of the time (and a fraction of the cost) on a D7000 or a 5DMkII...

[rainer_v]

i am pretty sure it will not stitch the raw files, in case the stitching is done automatically ,- as the sony doesnt do it either. it would be a great feature if the raws would be stitched together !! but if the outcome will be a jpeg its more or less worthless for most serious purposes.

and i insist: sure its great that the red puts out usefull stand images with print quality, but i dont see them closer to the hiend or medium end still cameras than i see the video output from these to the video files of the "real" video cams as the red. this video files from canon and its brothers/ sisters are usuable in certain commercial purposes as well. the gap between this two worlds is getting closer, but its still there from both sides and i.nm.o. it will also be there for some years. and i dont care what will be in 4 years ..... thats an eternity in our fast times and so much can change ....

[ChristopherBarrett]

Just for fun... here is a project I shot last year.  One image is form the Arca / P65+ the other is from my Red One.  These are from my blog, so they're not huge but it's fun to compare.  The Red Workflow is very similar to Capture 1 Pro and it doesn't take any more effort to get an exceptional still out of it.  Personally, I love all my toys....ahem... TOOLS!





I didn't happen to color these to match, I hadn't planned on making a comparison... and am quite pleased with both.

CB

[BernardLanguillier]

Very easy: Lik makes $$zagillions$$, so proof is there are lots who like it...

Oh yes... but if making money were the measure of all things I would be working as a trader or selling missiles to developing countries.

Since this is a RED ad, I would have expected them to come up with something that potential RED customers - like people at LL - would find impressive technically and aesthetically. I may not be representative, but it fails on both accounts for me.

Cheers,
Bernard

[bcooter]

Yair,

With all due respect are you thinking like an image creator's client, or like a photographer or a camera rep?

All are a different thought process, but in imagery for commerce the only real final decision maker is the client, or the client's, client's, client's boss.

Because if your thinking like a client on a production that is shot in both stills and motion, here's the process.

They first select their favorite stills and corresponding motion clips for the creative brief.

Then we go into post production, editing and retouching on stills, editing and coloring and effects for motion.

Somewhere in this process someone scrubs through a quicktime movie, stops it and sends a screen shot.  In our case usually 2 dozen screen shots and says "this would make a great still, can you make a still image of it?".

My answer with the RED is yes and though some of the images are technically spectacular, some are more challenged, the client is happy, actually over the moon happy because they had more selection.

(I'd show them but due to extended deadlines, most of the large projects we've shot this year are still under embargo).

Now do I think this image shown by Mr. Lik is better than a phase back?  I don't know, don't really care, but I am positive that this and the many dozens of still images we have produced from our REDs have been accepted and used.

I'm not subscribing to the one camera fits all mentality because that doesn't happen in any medium,  but I am absolutely positive the more you can give a client the better your business is.

I think you know this.

I've shown this image before, and it's from a motion session, shot on the RED, using an old Nikon 17mm zoom resting on a jacket.



Shooting reverses, sound, tight, wide, the total session ran about an hour.

Is the image perfect . . .? . . . no not in technical terms that this section of the forum holds in high esteem, but is it useable and sale-able, yes, as it's run as a double page print ad.

Now to be honest, I don't know why I even mention this, because it doesn't put a penny in my pocket. With Red I pay list, as with most of my equipment and whether another "still" photographer ever shoots any camera that fires more than 2 frames a second means nothing to me, but that doesn't mean that to some extent, still and motion imagery has converged.

And Yair, you also know that working 6 hours in post, on any still image of importance, is not unheard of, in fact in today's world it's almost a standard.

Rainier, yes you can pull stills from a 5d2, but the difference in a motion capture to still from a 5d2, vs a 5d2 still capture is quite different.

http://ishotit.com/dsmc1.jpg

The Red file I posted above in it's large form is close to the 5d2 still image.  Not exact but very close.



IMO

BC

[georgl]

For some filmmakers (especially when a small size is crucial) this is a valid tool, for a photographer the Epic is a 2kg heavy 15MP APS-C (well, odd ratio, a little bit wider) mirrorless camera with a compression before debayering - some kind in-between JPG and RAW.
The sensor is current CMOS standard, it offers low-noise even at 800ASA and the dynamic range is quite high (I've never seen proper independent tests that go far beyond 11 stops, which is quite good
The HDRx is nothing but a software-function with two separate exposures with the known issues for moving objects.
The ALEXA is a different beast with large photosites and patented dual-gain sensor with over 14 stops of actual dynamic range - which is unheard of. But it's not a photographers tool, either. It's large and heavy (because it's build to work reliable under tough film-set conditions) and has about 6MP on it's sensor. ARRI would never market it as a photographers tool or IQ180-alternative, that's just stupid and therefore they don't even try it.

[yaya]

I (think) I understand everything you say james but the claim behind the RED Vegas panorama is hardly backed up and if I could make the same (or perhaps better) image with another tool and do it faster and easier, for me as the image maker and for the client and the client's boss as those who pay me then I think I would pick that other tool

If such an image takes 6 hours to create (maybe less, maybe more, we'll never know...) and another camera takes 30 minutes then for me the choice is quite clear, especially if the client's requirement was, let's say, a 10' 360dpi print (which is not unheard of these days)

BUT if the RED hardware+software can make this kind of seamless panos straight out of the footage and spit out a decent size file in some sort of RAW or high res, high bit-depth format then that would be a big step forward for many applications...

[fredjeang]

I think that this mysterious "Alchemy" that we'll know soon about, has nothing to do with stiching because it's a slider.

IMO, this image is actually a night shot, then using Alchemy you recuperate color information without degrade. Something like that. I'm speculating of course but what else could it be as a slider?
And if this is indeed a night-shot, Red engineers have done a bomb.

Back to this thread, I think that a lot are not seeing the all picture honestly and take things outside the contexts.

Again, this has nothing to do with this pic in itself, if it switched or not and how many hours it took to deliver the printing, or if this image is horror or not.

It has to do,  that it comes from a single device, basically a high performance video camera, that has the size of a MF camera, on wich you can mount many glasses, that shoot raw-video with high dynamic range on par with MF, if not better, and capable of deliver exploitable still imagery, and with some trickings, do big enlargements too.
All that, in a price package somewhere reasonable if you think about the fact that we are talking about high-end video gear.
And a free developper that everybody can download on Mac or Win that allows to do some rough-cut and generate stills, do a prim grade, very much like a capture one software.

Also, R3D files are generational-friendly: every enhancement done in the new generation are compatible with older ones, in short, no new fancy codecs that nothing reads. (and that is something the competion is not able to understand at the moment.)

[eronald]

Yair, Old Branch,

Should we congratulate you on being promoted up the tree?
Do you manage the Mamiya/Leaf product packages in Japan too?

Edmund

BTW: Old french child's quiz:
Q- How does an elephant climb down from a tree?
A- It holds on for a leaf and waits for autumn.
Follow-up
Q and why are crocodiles so flat? ...