FAQ: Canon EOS 450D / Digital Rebel XSi Volume III


18.06.2009 11:31:00

12 and 14-bit ADC is and it has nothing to do with the matrix, and it was there, in your opinion, "pulls". Not to be confused with a soft warm.
 

18.06.2009 11:31:00

on what basis? facts, facts

Facts I quoted above. The conclusion of the computational experiment - or ADC Canon EOS 450D 14-bit "ny or programmers Canon big jokers and recorded two least significant bits of RAW pseudo-random sequence.

To refute the second hypothesis is sufficient to produce intentionally overexposed frame (all three channels overexposed) and drive through the same program. There will be a significant distortion of the residue in a single number (corresponding to the value of the lower two bits of saturation). I would be grateful to anyone who will do it, or will send me a picture.

NB! When making test overexposed frame should keep the shutter speed is less than 1 second exactly (preferably not more than 1/15, for the warranty) - 1 second is usually included noise reduction software.


And the real difference is insignificant in the DD, but not because of the fact that between 12 and 14-bit there is no difference, but because real matrix does not pull as much

Sorry, DD sensor and ADC bit depth - it's about various things.

18.06.2009 11:33:00

honestly ADC outputs 14 bits, but the lower 2 bits are filled with noise and useful information shall not be

18.06.2009 11:37:00

now looked several reviews in the sources that I trust - always write something like "14-bit processing, that should, at least in theory, give a higher dynamic range."

Not everything you read on the Internet, can be trusted. In this case the quoted statement about the dynamic range should not believe - DD bit depth ADCs and meter values ​​tend skorellirovannye but not directly dependent from each other. Increasing bit ADC may be due to an increase DD device and desire and goodwill developers measuring and digital complex to give the user access to more accurate data. Or it may not be.


honestly ADC outputs 14 bits, but the lower 2 bits are filled with noise and useful information shall not be

Where information about the noise?


18.06.2009 11:41:00

And you found this preskorbno fact?

18.06.2009 11:49:00

enough to make intentionally overexposed frame
entire frame has to be white? Or not necessarily?

18.06.2009 12:15:00

entire frame has to be white? Or not necessarily?

whole. I would recommend:
a) Rasfokus, opening the aperture, ISO100;
b) Partial metering at the ceiling, a gray card or sky begin to lengthen the shutter speed:
a) +2 EV vignetting (with reserve);
g) +3 EV setting on the light meter;
d) +1 EV exposure compensation occult (details here );
e) +1 EV sensitivity to different channels;
w) +1 EV in reserve.

Total shutter speed will have to be lengthened by 8 foot (256 times). If he measured the exposure meter 1/1024, set in M ​​mode shutter speed of 1/4. Be what you need.

Control reminder! Shutter speed should be kept shorter than 1 second for sure. I would even recommend shortened to 1/4 or shorter (t. E., Need a lot of light).

18.06.2009 12:52:00

honestly ADC outputs 14 bits, but the lower 2 bits are filled with noise and useful information shall not be
where did you get that noise?
you which camera is not 12 bit hour? you are so defending uselessness 14 bits


shutter speed should be kept shorter than 1 second
exactly what you're talking about denoising? dark frame subtraction, or what? Well this is equal to


18.06.2009 13:20:00
I
450D. Otherwise, what would I do in this thread. I understand that everyone sandpiper praises its swamp.
Had the opportunity to compare with the 50D and 5D (not on the same pictures though) - the difference is significant. Besides that
5D exactly 12 bit.

My deep IMHO, that in this case we limit it analog matrix
14 bit - it's good, if applied to them less noisy matrix or higher DD - would be even better.

As for the experiment. It would be interesting to see the raw one and the same frame c 400 and 450.
plot - something like the gradient goes beyond DD.
can be assessed if any difference in the dynamic range of the matrix and the difference in the transfer of semitones (or the difference will be clogged with noise).

18.06.2009 13:36:00

As for the experiment. It would be interesting to see the raw one and the same frame c 400 and 450.
supported. Does anyone have the opportunity to do this?

18.06.2009 13:39:00

50d in a noisy matrix, so you know (pixel by pixel)
but you can see the difference
I do not either because the number 50 in less than 450?

and 5d did so, has a higher DD because of its pixels (senseley), it has 14-bit ADC would show better than 450d




there are only 450 and 500


18.06.2009 13:45:00


Can a stupid question? And why do it?
Where two extra bits per color to be hammered noise I just do not understand, okay if the matrix at 400 and 450 were the same, but different percent could unsure, as well - there is no doubt. From overexposure to stop details of stretch easily, I think it is due to the 14-bit

18.06.2009 13:48:00

and I'll say, of the lights 450d provides more information than the 400d
matrix allows it or ADC, it does not matter in the end
greater DD 450d there

why two additional bits will only noise? of what it should be?


18.06.2009 13:53:00

you about some denoising? dark frame subtraction, or what? Well this is equal

Pro that turns at speeds of 1 second or longer. In some cells. Prior to recording in RAW. In the next topic Research noise SLR digital cameras in the topic starter has links to the plots of the measured noise dark frame depending on age. Some cameras after 1 second observed decline in the level of noise, which indicates the presence of noise reduction. For example, look at the schedule here Nikon D70.

dark frame subtraction - it's about something else.

18.06.2009 14:22:00

why two additional bits will only noise? of what it should be?
LSB is always the most noisy.

18.06.2009 14:31:00
Guys, need urgent help, what are the symptoms unopened 450D?
Beru in the internet. Box on each side of the red, the rest - white, looks like white box, so as not to lopuhnutsya?

18.06.2009 14:33:00

I asked why a person believes that there is only noise, and no more noise than other discharges


vaytboks all white, no pictures of the device and the red


while kenona have a squelch?
know that Sony and Nikon have just


18.06.2009 14:48:00
Here man tried to compare the DD 450 and 400. There is a difference, and it is minimal.
http: // ecamera. tkat. ru /? mod = articles & amp; act = full & am ... = 1053 & amp; page = 17
But too little information in the article.


why two additional bits will only noise? of what it should be?
because the matrix itself is noisy. And at high ISO noise amplitude of this many times more than the resolution of the ADC.
At a minimum you need to watch the ISO test shots - there is a difference in the transfer of parts or their score as noise.
http: // www. the-digital-picture. com / Reviews / Canon-EOS ... amera-Review. aspx


as well - there is no doubt. From overexposure to stop details of stretch easily,
12 bits of overexposure 1-2 foot parts are drawn as easily. A clear example is 5D.

Here it is necessary to understand what these are spent 2 extra bits:
1. an increase in DD. Fotkat line stretching from the bright lamp in the darkness. Twist in RAW exposure compensation, and look to what division we can pull something.
2. a more accurate dithering. That is DD remains, just broken into smaller segments. Here is our main enemy - the noise.
Again, look at whether we can distinguish 450 items that were not visible in 400.

Regarding n. 2 in the test www. the-digital-picture. com an example, it is true between the 450d, 50d and 5d.
5d with its 12 bits of tears due to the large size of the matrix and thus meshshe noise.
450d and 50d on Dithering about equal, even a little better here 450d rights

18.06.2009 14:55:00

while kenona have a squelch?
know that Sony and Nikon have just


If you are not too lazy to look the same topic summary picture in the first posting, you already would know that, at least in the 40D is, and 20D, too, and to 5D, and 400D. And there has not been investigated 450D (450D neither the owner of the 20th offline data not provided). In the absence of better information to suggest that there are surprises.

18.06.2009 15:05:00

so it is clear that at high ISO noise
and on minimlnom, I would not be sure that there is only noise
otherwise used did not do 14-bit


well, not looked
clear the remaining time there is such a canon, and then there are probably 450 (although it is different 14 bits of his, and 100% is not known whether or not there shumodava long exposure)
ready to provide you test rabbis write conditions
as I understood, the conditions are the same as the bit you named above, only put a few excerpts, much less one with and about?
but at the same time ensure equal brightness of the subject?

18.06.2009 15:08:00
Thank
, then he is not. Boom to take!
And the machine is not sealed in plastic? How to define "fresh"?

18.06.2009 15:09:00

I bought in an Internet store, and the box was opened (I went to their office and took the machine)
whether they are not sealed, whether opened before buying my
but the frame counter on my machine only my footage

18.06.2009 15:15:00

ready to provide you test rabbis write conditions

Thank you. However, I do not need. I'm not going to buy the 450D

upd. EMNIP, filmed at different ISO disk loaded lid closed (paranoia!) Lens at shutter speeds of 10 sec, 1 sec, 1.8 sec, 1/60 sec and 1/200 sec. In the subject method is described, given the link to the software, you can do an experiment at home.

18.06.2009 15:30:00

closed lid disk loaded (paranoia!)
weird
can, and try to himself

18.06.2009 16:25:00
In general, after long readings. and reflection.
decided to purchase a future date Canon 450D KIT
But I can not understand why some stores, there are different positions of the SLR with different prices, description Kotya no oltichayutsya.
or I just do not see?

18.06.2009 16:33:00

in the same store?
give an example
can, complete with memory cards, bags and so on. D.
lens with or without
and objects can be different
 

18.06.2009 16:43:00
Yes, I am familiar here suggested that Lenses can be different.
and watched so many places.
in the last remembered pleer. ru
though for some reason the reviews about this Magazev not)

18.06.2009 16:45:00

In general the fact that the box is still open, and well they should. package open documents t. To. need warranty as to put down. That thought, as yet to be determined externally, that bird is not a test and reject "non-virtual" customers. Very far from the capital, do not touch.
assured that there are unopened boxes, photo sent. Store anything before like myself not compromised, so hopefully all will be ok.

over the counter staff - at a clear, there programm

18.06.2009 16:52:00

Normal boutiques, and the prices are good, and do not guarantee left. There I took a phot

18.06.2009 17:03:00
By the way, about the price of 450D. I live in the Volga region in the city with a population of 100 thousand. One person, the owner of a computer store, said he could bring 450D with kit lens for 22,000 rubles (assuming 100% prepayment). Says that everything is legal.

Is it real? What to look out for in the first place and how to protect yourself from buying low-quality / not a new product?

18.06.2009 17:08:00

officially shipped without a lens, or one lens (kit) - 18-55 is (note - the letters is REQUIRED)
like there is still a double set of 18-55 is and 55-250 is - but I live in this kit only saw have 1000d
if sold to another lens, it is no longer the official kit, and the warranty on the lens may be missing or what is left (from the store, but not from kenona)
recently to 1000d often shoved the old lens 18-55 II (although there is also the official kit with this lens), and possibly also to the 450d already shoved his
need to check here

old ef-s 18-55 II

but new ef-s 18-55 is


18.06.2009 17:11:00

while kenona have a squelch?
and described in the documentation.

18.06.2009 17:18:00

here talking about shumodava in raw, before writing to the card and the
that high iso denoising of custom settings, like, only for acts zhpega

18.06.2009 17:20:00


450D with kit lens for 22,000 rubles
too cheap for the new 450D kit, do not believe. 23. at least 5-24 thousand somewhere, and then in Moscow. Is that if someone in the warehouse from last year stray .

18.06.2009 17:21:00

weird - the diaphragm cover with the lid closed
or there in the article about this is? (In English, I do not read so well, that immediately grab)

18.06.2009 17:23:00

weird - the diaphragm cover with the lid closed

And you about it. Well I wrote above - paranoia

upd. There's the basic idea - to take a picture of nothing. E., No photon is not absorbed by the sensor should during exposure. Then all the noise, which will be - their own, "instrument".

18.06.2009 17:30:00

too cheap for the new 450D kit, do not believe. 23. at least 5-24 thousand somewhere, and then in Moscow. Is that if someone in the warehouse from last year was overlooked.
And as an option to bring from abroad?

18.06.2009 17:41:00


Well, that's just "gray" and may have a place for the 22 th. But if you do not need a guarantee.

18.06.2009 17:46:00

recently to 1000d often shoved the old lens 18-55 II
Incidentally, this is probably not amateur sellers, these are the sets 1000D + old whale watching now abroad in different stores. 450D are all the new whale.

18.06.2009 18:45:00

not, in my case, it was amateur, as it is written in a price 18-55 is
and the box, most likely from the carcass

18.06.2009 19:07:00
I
AMD-took his photo in the middle of May, "pereschupal" 3 samples, with a focus in all everything was fine, on the 1st on the screen was broken pixel, and on a matrix of all were hot / broken, select the sample with 3 hot , visible only at about 2 second exposure, on the whole thing has spent 2 hours (store employees kept a stiff upper lip, I thank them ) No box was not sealed, and then all sprashival- at all so it was. Warranty Card of a new model, gray such online Canon "and it is. In general, satisfied with the purchase

18.06.2009 20:04:00
quote:
:

here talking about shumodava in raw, before writing to the card
I talking about?

18.06.2009 20:15:00

is the one that the high iso?
he chroma noise in zhpege presses

18.06.2009 20:19:00
quote:
:

is the one that the high iso?
he chroma noise in zhpege presses
I'm talking about the dark frame subtraction for long-term (more than 1 sec.) Shutter

18.06.2009 20:22:00

then I will immediately asked - whether it's dark subtraction frame?
says - not just
automatic noise reduction at slow shutter speeds in raw
all previous devices, it is written in a thread on noise, was probably
and 450d have

18.06.2009 20:41:00
quote:
:

then I will immediately asked - whether it's dark subtraction frame?
says - no
e. Type 3 shumodava?

18.06.2009 21:12:00

such as, for tests, it turns out noise reduction in raw extracts from the aspect ratio, after which it is enough to slow shutter speeds, there is a decrease in the noise level, while up to this point, the noise only increased
FAQ: Canon EOS 450D / Digital Rebel XSi Volume III. # 2826

FAQ: Canon EOS 450D / Digital Rebel XSi Volume III. # 2821


19.06.2009 5:56:00


Checked himself, it turns out, there is no "third type shumodava." Attachey graphics dark noise at ISO 100 ... 1600 and exposures in the range of 1/8000 ... 8 seconds in increments of feet on the example of the camera Canon EOS 40D. It is clear that additional noise reduction at slow shutter speeds not. Growth with increasing noise exposure due to the influence of thermal noise (on a linear scale quite clearly seen from the linear growth pattern for long exposures).

I beg your pardon, it seems, did not understand until the end of the procedure of "and.
 

19.06.2009 8:53:00

ie no noise reduction can have no
vrgrin dark frame subtracted?

FAQ: Canon EOS 450D / Digital Rebel XSi Volume III

info@www.about-digital-photo.com